Please enable JavaScript.

Coggle requires JavaScript to display documents.

Computer Science ab initio, HTML, fun intersection of probability theory,…

- - - - Browsers display webpages
        
        URL = Uniform Resource Locators. Web browser addresses of internet pages and files/
        
        3 parts to a URL to address a page or file
        
        Protocol
        
        Most web pages use http or https, but there are others. Hypertext Transfer Protocol +/ Secure. HTTP is Data communication standard of web pages. Addition of ~secure~ signifies additional layer of encryption
        
        Portion ending in //:
        
        Communication protocol = system of rules allowing 2+ entities of a communications system to transmit information via any kind of variation of a physical quantity. The protocol defines the rules, syntax, semantics and synchronization of communication and possible error recovery methods. Protocols may be implemented by hardware, software, or a combination of both.
        
        Protocol types include
        
        3 more items...
        
        Host
        
        Or top-level domain
        
        Frequently ends in .com, .net, .edu, .org (though many others)
        
        Filename
      - The web address is an internet address of a website, file or document (which all have unique web addresses being connected to the internet)
  - - - Department of Defence of USA developed a network called Advanced Research Projects Agency Network - ARPANET
- - - - WAN
        
        Extends over large geographical distance
        
        Often established with leased telecommunication circuits
        
        Wide-area networks
      - LAN
        
        Interconnects computers within a limited area eg 1 building
        
        VLAN
        
        Subnetwork group of devices on one or more LANs configured to communicate as if
        
        Virtual local area network
        
        Local area networks
        
        Most LANs use the Ethernet to connect together
        
        Ethernet
        
        Communication standard
      - SAN
        
        Storage area network
    - - Hubs can be used for...
        
        Historically, the main reason for purchasing hubs rather than switches was their price. Today few hubs are produced, because they are not an effective solution for building networks.
        
        working on Physical Layer of OSI.
        
        inserting a protocol analyzer into a network connection
        
        A hub is an alternative to a network tap or port mirroring.
        
        preventing a "loop" by accidentally connecting two ports together, which might be a big problem when switches are used.
      - Switch(ing) hub
        
        Reads the destination address of each packet, then forwards the packet to the just correct port
        
        lower latency than ethernet hub
      - Ethernet hub
        
        Acts like a repeater; all data arriving to one port is sent to all other ports (without finding destination using IP address). All devices in this network will get all the data.
        
        does not operate ethernet traffic
        
        Simple, configurable electronic device that runs a computer network
        
        Another type is the passive/intelligent hub
        
        Device monitors traffic going through the hub and all ports
        
        Can be configured (aka manageable hubs)
        
        Hub described here does not look at individual packets - it copies data to all ports and is used as connection between LAN segments
        
        Often has LED signals to show which ports are connected and if hub is working
        
        A collision happens when 2+ devices on the same network try to send packets simultaneously
        
        Then, all devices have to go through a routine to resolve the conflict. This process is set in the CSMA/CD protocol (Ethernet Carrier Sense Multiple Access with Collision Detection)
    - - Transmission Control Protocol/Internet Protocol
- - - - Cache memory
      - CU
        
        Instructs memory, arithmetic/logic unit, input + output devices how to respond to a program's instructions
        
        Directs operation of the CPU + controls transfer of instructions and data + coordinates parts
        
        Decodes opcodes for different operations in a CPU
        
        https://computersciencewiki.org/index.php/Control_unit_(CU)
        
        A control unit coordinates how data moves around a CPU
        
        Obtains data/instructions from memory --> interprets/decodes instructions into commands/signals
    - - ROM
        
        Non-volatile
        
        Used to store programs permanently
        
        Read Only Memory
        
        Eg start-up or boot instructions
        
        Non-modifiable- cannot be written to (almost always; if so, then its once and never again) or taken away from
        
        Usually used for firmware in CPU, Graphics card, hard disks etc
        
        Many are replaced by Flash modules
        
        BIOS
        
        Allows easier updating
        
        ROM replacement
        
        May still be used in devices like keyboards - not designed for firmware updates
        
        Stores code for interrupter
      - RAM
        
        Writable --> Can be read, written and erased any number of times
        
        Volatile (though non-volatile has been developed) --> loses contents when power is removed
        
        Stores programs and data being used by the CPU in real time
        
        Random Access Memory
        
        Random-access = ability to access an arbitrary element of a sequence in equal time
        
        Device allows data items to be read or written in (almost) the same amount of time - irrespective of the physical location location of data inside the memory (unlike other direct-access storage media eg hard disks w/ mech lim)
        
        Engineering
        
        (usually) more than one bit of storage is accessed by the same address. RAM devices (often) have multiple data lines. 8/16-bit etc devices.
        
        Currently, components are IC (integrated circuit) chips with MOS (metal-oxide semiconductor) memory cells.
        
        RAM contains multiplexing and demultiplexing circuitry to connect data lines to addressed storage for reading or writing the entry
        
        Each instruction and piece of data within has a unique address
      - Time between primary memory and CPU is 7 to 10 nanoseconds
  - - - Manage hardware interfaces
      - Allocating storage
      - Swapping between programs
      - Memory Management
        
        Less common pages not loaded. When full, pages not being used (or calculates the best one to remove) are swapped for those that are needed.
        
        Memory map lists what blocks have been allocated to each task.
        
        Only necessary pages of a program are loaded to avoid filling too quickly
        
        Virtual memory is used when program size > available memory.
        
        Allowing more space for other applications.
        
        May slow operation
        
        Only loads pages from storage when needed.
        
        Sections of secondary storage treated as
        RAM or hold a central block of code in RAM
        
        Memory is split into pages
        
        Aims to allocate so multiple programs can run
        
        Protects processes from each other
        
        Enables sharing of memory space
        
        Helps provide a satisfactory and optimized level of performance
        
        https://www.geeksforgeeks.org/memory-management-in-operating-system/#:~:text=Memory%20management%20is%20a%20method,achieve%20efficient%20utilization%20of%20memory.
      - Network Management
      - File Management
        
        Hierarchical system to store files using folders or directories.
      - Interrupts - Inputs/Output. Copes with errors
      - Input/Output (I/O) Management
        
        Most I/O devices operate operate much slower than the CPU, so a queue is created (sequence or priority) to complete requests/processes
        
        Control driver programs load with the OS to detect I/O devices, allow device to communicate with the OS
        
        An I/O buffer sits between the CPU and slower devices, so the processor does not need to wait for a slow response.
        
        Manages peripherals: any input and output devices connected to the CPU
      - Processor Scheduling
        
        If a task is inactive, time slice
        might be passed to the next task.
        
        Illusion of multitasking
        
        Multitasking also occurs with multiple cores
        
        If time slices are too small, the data may not save - leading to corruption
        
        Round-robin giving equal time.
        
        Processing time divided up between tasks
  - - - An algorithm can generally be performed over lots of time(++ clock cycles) with few hardware resources or over fewer clock cycles using a large number of ALUs, registers and memory.
      - Execution speed v hardware complexity tradeoff --> Variety of hardware microarchitectures can be generated from one algorithmic description through an HLS compiler
  - - - Assembly Code
        
        Not portable, 1-1 relationship
        
        Quick execution
        
        Uses abbreviated codes for instructions
        
        Compact due to optimisation
        
        One-to-one relationship with machine code
        
        High level manipulation of processor registers
        
        Used in real time applications (vital fast response)
        
        Also used in device drivers, embedded systems (due to direct hardware manipulation)
      - Byte Code
        
        Requires a virtual machine to turn byte code into platform specific machine code
        
        example Java
        
        Java IDE compiles bytecode .class file
        
        Bytecode can be sent to any machine
        
        .java files can be written and compiled on any machine using the Java IDE
        
        As long as target machine has a Java Intermediate Virtual Machine installed, bytecode is converted into platform specific machine code.
        
        Overhead, but allows any Java code to be run on any machine with the correct VM installed
        
        Intermediate language between a high level language and machine code
      - Machine Code
        
        Written in binary based on CPUs instruction set
        
        Every language is translated to machine code
        
        lowest level of code (0s and 1s)
        
        High risk of input errors
        
        Not portable to different architectures
        
        Hard to find errors
        
        long:(
        
        Used where fast execution or direct processor access is needed
      - Assembler
        
        C
        
        Turns assembly code into machine code
      - eg
    - - Object-Oriented
        
        Runs once object has been interacted with
        
        Encapsulates data
        
        Instructions held in objects
        
        Bottom up design
        
        Uses classes:
        
        Functions
      - Event Driven
        
        Event/trigger functions
        
        Subroutines
        
        Time driven
        
        GUIs, user-input based action
        
        Program flow is determined by events
        
        Sensors
        
        Messages
        
        User input
      - Python
      - Algorithms
      - Procedural (Imperative)
        
        Subroutines
        
        Open data sharing
        
        Process focused
        
        High-level synthesis (HLS)
        
        Abstract behavioral specification of a digital system --> register-transfer level (RTL) structure realizing given behavior
        
        eg. C, C++,MATLAB
        
        Automated design process
        
        Synthesis begins with high-level specification of the problem decoupling behavior from low-level programming
        
        Aka C synthesis, electronic system-level (ESL) synthesis, algorithmic synthesis or behavioral synthesis
        
        Code is analyzed
        
        Process activities
        
        Library processing
        
        Resource allocation
        
        Control/Dataflow analysis
        
        Scheduling
        
        Algorithm optimization
        
        Functional unit binding
        
        Lexical Processing
        
        Register binding
        
        Output processing
        
        Input Rebundling
        
        Factors influencing HLS
        
        Low-level timing constraints
        
        Iteration
        
        Loop
        
        Memory
        
        Interface
        
        Hierarchy
        
        Computer follows a list of instructions every time
        
        Command Line
        
        Used in first HLLs, CLIs and basic programs
      - Declarative
        
        Non-destructive assignment
        
        Non-crucial order of execution
        
        Describes what the program should accomplish rather than how
        
        Explicit data structures
        
        Used in logic programming, fact-based results
        
        Used in AI
      - C
      - C++
      - Natural language elements
    - - cross-compiler
      - bootstrap compiler
      - Usually for programs translating source code from a high-level programming language to a lower level language to create an executable program
      - Operations
        
        Parsing
        
        Syntax/Semantic analysis
        
        Lexical analysis
        
        Preprocessing
        
        Some or all of the below processes
        
        Often called phases
        
        Input program conversion to an intermediate representation
        
        Code optimization
        
        Code generation
      - Program that converts entire source code into object code before execution. Produces full error report once compiled.
      - Implementing operations/phases as modular components promotes efficient design and correctness of transforming source input to target input
    - - Each statement is converted then immediately executed
      - Computer software that transforms then executes the indicated operations
- - - - Binary Errors
        
        Underflow and Overflow
        
        Will cause program to crash or erroneous results
        
        Overflow when a number is too big to be represented by the number of bits allocated - 1 (00000001) and 127 (01111111) would produce 128 (10000000) but in two's complement, -128
        
        Underflow = number is too small to be represented with the amount of bits allocated. Eg 1/32 representation impossible using 8-bit fixed point numbers, as smallest that can be represented is 1/16
        
        Rounding errors
        
        Absolute rounding errors
        
        Relative rounding errors
      - Base 2 (0 and 1)
      - Error Checking and Correction
        
        Checksums
        
        Returns value; if equal to source value, then transmission was successful
        
        Result of a cryptographic hash function
        
        Check Digits
        
        Modulo-11 -
        
        Eg 23045 is sent as 230456 so check digit is added and then modded until the value reaches 0.
        
        Digit added onto the end of binary data to check its accurate
        
        If calculation performed on receiving end does not match, an error message is created.
        
        https://www.loc.gov/issn/check.html
        
        Majority Voting
        
        eg 10 sent as [4] = 11110000
        
        Each bit in a binary code is sent a set number of times, then checking if the binary digit is the same each time
        
        Invalid transmission of 10 as [4] could be 11010001
        
        Parity Bits
        
        In a byte, 7 bits are sent + the additional parity bit
        
        Even Parity: added to make total number of 1s an even number. This is checked on the receiving end.
        
        Additional bit added onto a binary string to ensure total number of 1s/0s even or odd.
        
        Odd parity: total 1s an odd number, also checked on receiving end.
        
        Parity bits can be generated by XOR
        
        USE: fault tolerance
    - - Base 10 (0 --> 9)
    - - Shorthand for binary - 2 hex digits = byte
      - Less digits than binary + easier comprehension than binary strings
      - Base 16 (0 --> F)
      - Used in HTML hex color codes, MAC addresses, memory dumps, error messages
  - - - Kilobyte (k) = 10^3, megabyte (M) = 10^6, gigabyte (G) = 10^9, terabyte (T) = 10^12
      - Units - Binary increments in 1024, Decimal units increment in 1000
      - kibibyte (Ki) = 2^10, mebibyte (Mi) = 2^20, gibibyte (Gi) = 2^30, tebibyte (Ti) = 2^40
      - Byte = group of 8 bits
      - Bit Patterns
        
        Digital = discrete (discontinuous) values to represent values
        
        String of binary represents high/low electrical signal --> turned into digital signal --> appropriate form of data
        
        Analog = uses continuously values which produces a continuously changing waveform
        
        Can represent binary digit/signed binary digit, character code, pixel of a picture, sound note, length of a line, volume of a note
        
        ADC = Analog to digital converter turns real world continuously changing waveforms into digital signals. Computers use ADC to interpret analog signals produced by sensors
        
        Samples analog signals at set time values -> records amplitude -> quantization rounds amplitude up or down so value fits in bit resolution
        
        DAC = Digital signal to analog waveform that devices can use. Eg speakers turn a digital audio signal into an analog signal for speakers
        
        Sound
        
        Analog signals vary continuously --> readings must be taken at fixed intervals
        
        Sample resolution = amount of intervals of amplitude recorded
        
        Analog sound waves must be samples to turn them into digital waves.
        
        Sampling rate = number of samples taken per second
        
        Nyquist's Theorem
        
        Sampling rate should be at least double the highest frequency desired
        
        Ensuures complete coverage of peaks and troughs in the analog signal, allowing faithful sound reproduction
        
        Upper frequency for humans is 20,000Hz - analog frequency must be sampled at 40,000Hz
        
        filesize = frequency (of sample rate) x bitdepth x duration
        
        MIDI
        
        Instead of direct sound representation, it transmits data to control how sounds are produced
        
        Smaller file size than pure sound file (loads faster), fully editable/customizable, supports huge range of instruments, high quality and authentic sound reproduction
        
        Protocol for recording and playing back music; controlling devices
        
        Event messages = control the properties of a sound in a file, recorded in binary - communicate between a MIDI device and the processor
        
        Note length/volume/velocity/pitch
        
        Effects to apply to notes
        
        When to play/stop a note
        
        Timing notes to play with other sounds/notes or with other MIDI devices
      - Bit = fundamental unit of info, 0 (off) or 1 (on)
    - - Binary Number Systems
        
        Signed binary = binary that represents a positive or negative whole number
        
        Adding unsigned binary adds the two numbers together in binary
        
        Signed number representations are required to encode negative numbers in binary number systems. https://en.wikipedia.org/wiki/Signed_number_representations#Sign%E2%80%93magnitude
        
        ones' complement
        
        sign-magnitude
        
        two's complement
        
        Most significant bit is negative
        
        Two’s complement range
        
        Maximum representable value (2n-1-1)
        
        Minimum representable value = (-2n-1)
        
        Converting to denary
        
        An 8-bit two’s complement number has -128 as the most
        
        significant bit.
        
        Write out denary equivalents (with sufficient columns
        
        (doubles each time) for the length of the binary string)
        
        Enter binary string in (eg. 10011100)
        
        Add up values (eg. = -100)
        
        Represents negative integers (signed binary)
        
        Converting from denary
        
        Flip the bits and add 1
        
        Flip all bits
        
        Write out the positive binary equivalent (eg. +102 = 01100110)
        
        Perform binary addition with 1 to get the answer
        
        Computers can only add binary {to perform the
        calculation 20 - 13 = 20 + (-13) - 20 = 00010100, -13 = 11110011}
        
        The answer can be checked by converting the final
        answer into denary (In this case there is an overflow bit handled separately to the calculation)
        
        Calculate range: using 8 bits, 256 numbers can be
        represented (-128 to 127). Arbitrary number of bits: 2n numbers can be represented. Unsigned binary: 2n-1 (as 0 is representable)
        
        offset binary
        
        Unsigned Binary = Binary that only represents positive whole numbers
        
        Min value (for n bits) = 0, Max value (n bits) = 2n - 1
        
        eg maximum number in 8 bits is (2^8 - 1) = 255
    - - Lossy
        
        Picture: grouping of like colors together in blocks so less data must be stored.
        
        Sound: perpetual coding removes frequencies less audible to humans and therefore less crucial.
        
        Video, picture, sound: Interframe compression keeps stationary pixels the same (continuing to show them) while only refreshing the pixels that have moved. Temporal/Spatial redundancy keeps identical pixel blocks within/between a frame. Cannot return to origina;
        
        Data degradation occurs - produces smaller files than lossless.
      - Lossless
        
        Dictionary Encoding: Stores language patters + common phrases in dictionary; these are referenced
        
        Needed for program code, text files, where data loss is disastrous
        
        Tokens are assigned to common text snippets, and then included in the compressed data
        
        Run Length Encoding: Sequences of the same repeated data are stored as a single data value and a count.
        
        Works on binary, pixel and character levels
        
        Eg: BBBBAAC = B4 A2 C1
        
        Data is exactly reproduced; none is lost
      - Reduces the number of bits required to represent data so files take up less space - often images and video, but can also be text files
  - - - Standard = 7-bit code - 128 permutations
      - Extended = 8-bit code - 256 permutations
      - American Standard Code for Information Interchange
      - Limitations are that 256 characters aren't enough, no multi-lingual support, web use requires universal coding system.
    - - Absorbed ASCII
      - Allows for more permutations and therefore more characters/languages/things
      - Standard binary coding system
      - Unicode Consortium defines 144,697 characters
        
        159 modern + historic scripts + symbols + emoji + non-visual control and formatting codes
    - - Important in
      - 16-bit code allows characters for 20+ countries, special characters, emojis
  - - - Substitution Ciphers replaces characters
        
        Caesar Cipher: One letter in an alphabet is mapper directly onto another in a shifted alphabet
        
        Frequency analysis: cracks it by assuming most common characters in cipher match with the most common letters in the origin language. One-to-one character relationship in such ciphers allow ease as different shifts can be tested. Further complexity added when polyalphabetic.
      - Computational Security
        
        If encryption method is computationally secure it may be crackable but note in a useful timescale
        
        Computational hardness = degree of difficulty in cracking a cipher
        
        How secure is data encryption?
      - Transposition Ciphers move characters to create an anagram
        
        Rail fence: message is split across several rows of a grid, where each letter is on the next row down and column. Encrypted message is formed by reading one row at a time
        
        Route: message fills a grid of set size, working down in columns. Encrypted message is read off row-by-row. For added security, transpose then substitute
        
        Vernam/Vignere/one-time pad Cipher: uses a one-time pad (key that is discarded after encrypting/decrypting). Key is chosen uniformly and randomly, so cipher text distributed uniformly
        
        Each character in the message and key's Baudot code are XORed together to produce another code
        
        New Baudot code turned into character, added to the encrypted string
        
        Random key of letters created; its the same length as the message
        
        Secure! Without the key, letters could be anything. Frequency analysis can't be used due to the randomly generated key. Regardless of time, it can't be cracked.
        
        Not perfectly secure as random generation is imperfect, key exchange might be intercepted, cannot authenticate sender/reciever.
        
        Only mathematically proved completely secure method as all other ciphers reply on computational security Given enough time and a long enough ciphertext, any other cipher can be broken.
    - - Cipher: non-decipherable method of producting ciphertext, Plaintext: comprehensible, Ciphertext: encrypted scrambled data
      - Key: data used to encrypt/decypt.
      - Link Title
      - Cipher cracking methods include assessing common techniques, dictionary attacks, reverse engineering, brute force
      - Encryption is the process of turning plaintext into scrambled ciphertext that can only be understood if decrypted.
    - - Binary code using crosses and dots
    - - Boolean logic operation comparing two input bits to generate one output bit
        
        If bits are the same, result is 0. If bits are different, result is 1
      - eXclusive OR